Game theory & social media marketing (4): Conclusions

The Royal Game of UR, Early Dynastic III, 2600BC, British Museum

[Part 4 of 4: Game theory & social media: Part 1Part 2, Part 3]

No, I’m no super lady, I don’t have no game whatsoever,
I put my high heels on and see how that goes, yeah
– Pauline, Sucker for love

Ask a mathematician why they like maths, and they will tell you that mathematics gives a definite yes or no. There is beauty in clarity. And, everyone likes to feel that they understand and have control over what is happening in their world. This feeling of certainty is reflected in the bottom two rows of Maslow’s hierarchy of needs: physiological and safety needs.

Tapping into fear and belonging

That said, we also love variety and surprise, which is the most popular information shared on social media. We crave new stimulus which is why we love games. We love the idea of chance or fortune transforming our lives for the better, and surely if we learn the rules, then we will succeed. And, that is why marketing has such a pull on us. Marketers tell us that we will have improved lives if we do/buy/or have what they are selling, and, marketers themselves will have improved lives too if we do/buy/or have what they are selling.

There are so many ways to market something, this link has 52 types of marketing strategies. The most effective, of course, aims at the bottom of Maslow’s hierarchy of needs – safety – which is why fear quite often drives news and coupled with specific instructions gives a compliant society.

Tapping into belonging is another way to market, which is why the connection economy and building friendship with your customers is gaining so much traction as a marketing strategy.

Modelling emotion and what-ifs

Modelling human emotion is impossible to do with game theory especially on social media, a fluid, still unknown, type of communication. We will never quite know who our audience is. We may target our demographic, but if they retweet or share something outside of that, then you never exactly know who is looking at your content, or how they will react to it. All game theory can do is offer interesting and potentially useful partial explanations to model a selection of what-ifs scenarios when employing different strategies.

In the last post (part 3), we looked at various game theory strategies from the aggressive to the altruistic, and saw that people generally behave like the people around them (hawk-dove) and that Kermit was in a bit of hurry to get together with his girl, which caused him to behave passive-aggressively, and probably not get what he wanted.

 Don’t be like Kermit

Game theory is a tool for social media marketing and the best application of it is recording trial and error attempts (with statistical significance) whilst using our emotional intelligence.

Be aware of your emotions and triggers (your personal competence) so you don’t get involved in a big wrangle either privately, which could damage a relationship, or publicly, which might be retweeted everywhere and could wreck your brand or reputation.  Even in the mathematics of game theory we need to understand other players moods and motives (social competence) and not assume anything. We need to ask for more clarification, so that when we do make a move, we do so with clarity and certainty that we are doing the right thing, and as any mathematician would tell you if you asked them, there is beauty in clarity for it gives us certainty and a sense of control, things which are harder to come by in our ever changing world.

Game theory & social media (3): What are you playing at?

Source: buzzfeed.com

[Part 3 of 4: Game theory & social media: Part 1Part 2, Part 4]

Whatever else anything is, it ought to begin with being personal – Kathleen Kelly, You’ve got mail (1998)

Kermit drinking his tea and throwing shade makes me laugh. However, I think we all understand his frustration. It seems that in business and personal relationships, people play games. We may not know why, and we may not know the rules. But as we saw in part 2, before we react, we might want to find out more: if a game is being played, which one, and if we want to play or not.

Games, payoffs, and winning

A game is normally defined as having two or more players, who have a choice of possible strategies to play which determine the outcome of a game. Each outcome has a payoff which is calculated numerically to represent its value. Usually, a player will want to get the biggest payoff possible in order to be certain of winning.

Dominance, saddles, and mixed strategies

Playing the strategy with the biggest payoff is known as the Dominance Strategy, and a rational player would never do otherwise, but it’s not always easy to identify which strategy is best.

So, players sometimes take a cautious approach which will guarantee a favourable result (also known as the Saddle Point Principle). Other times, there is no saddle point so players have to choose at random what strategy to play and hope for the best. They can calculate the probability of mixing up strategies and their chances of winning. If their probability skills are not great they can play experimentally and record their results 30 times (for statistical significance) to see which strategies work.

How does this work on social media? Well, no one knows how social media works so a trial and error approach whilst recording results can be useful. Luckily, Twitter and Facebook both provide services and stats to help.

Free will, utility, and Pareto’s principle

A major question is whether players have free will or not and whether their choices are predetermined based on who they are playing with and the circumstances in which the game takes place. This can depend on the amount of information players have available to them,  and as new information becomes available, they play a specific strategy, thus seeming as if they didn’t have free will at all.

Players assign numbers to describe the value of the outcomes (known in economics as utility theory) which they can use to guide themselves to the most valued outcome.

This is useful if we have a game where the winner doesn’t necessarily take all. If the players have interests which are not opposed and by cooperating the players can end up potentially with a win-win situation or at least a situation where everyone gains some benefits and the solution is not the worst outcome for everyone involved. This is known as the Pareto Principle.

On social media? Retweeting and sharing other’s businesses news is a nice way of ensuring everyone gains some benefits because with a potential market of 307 millions and there is enough of a market to go around for everyone to win-win and of course, reciprocate.

The Nash equilibrium

Taking this further is the Nash equilibrium which was named after John Nash, who proved that every two player game has one equalizing strategy (either pure or mixed) in each game. By looking at the equilibrium strategies of the other players, everyone plays to equalize. This is because, no player has anything to gain by changing only his or her own strategy, so it is win-win.

Are you chicken?

Ducks have been known share out the bread thrown to them so they all get some rather than one duck eating everything. This is known as the Hawk-Dove approach in game theory. When there is competition for a shared resource, players can choose either conciliation or conflict.

Research has shown that when a player is naturally a hawk (winner takes all) and plays amongst doves, then the player will adapt and cooperate. Conversely a dove amongst hawks will adapt too and turn into a fighter.

If there are two hawks playing each other the game is likely to go chicken, which is when both players will risk everything (known as mutually assured destruction in warfare) not to yield first.

We adapt very easily to what is going on around us, and on social media this is totally the same. In a 2014 study Pew Research Center found that people are less likely to share their honest opinions on social media, and will often only post opinions on Facebook with which they know their followers will agree – we like to conform.

The volunteer’s dilemma

In contrast, the volunteer’s dilemma is an altruistic approach where one person does the right thing for the benefit of everyone. For example, one meerkat will look out for predators, at the risk of getting eaten, whilst the rest of the meerkats look for food. And, we admire this too. We love a hero, a maverick, someone who is ready to stand up and be different.

The prisoner’s dilemma

But we hated to feel duped which is why the prisoner’s dilemma is one of the most popular game theories of all. Created by Albert W. Tucker in 1950, it is as follows:

Two prisoners are arrested for a joint crime and put in separate interrogation rooms. The district attorney sets out these rules:

  1. If one of them confesses and the other doesn’t, the confessor will be rewarded, the other receive a heavy sentence.
  2. If both confess each will get a light sentence. Which leads to the belief that:
  3. If neither confesses both will go free.

It is in each prisoner’s interest to confess (dominant strategy = 1) and if they both do that satisfies the Pareto principle (2).  However, if they both confess, they are worse off than if neither do (3).

The prisoner’s dilemma embodies the struggle between individual rationality and group rationality which Nigel Howard described as a metagame of a prisoner cooperating if and only if, they believe that the other prisoner will cooperate, if and only if, they believe that the first prisoner will cooperate. A mind boggling tit-for-tat. But, this is common on Twitter with those: Follow me, I will follow you back and constant following and unfollowing.

And, in any transaction we hate feeling like we have been had, that we were a chump, that we trusted when we shouldn’t have, which is why some people are so angry and like to retaliate. Anger feels better than feeling vulnerable does. But, great daring starts with vulnerability, the fear of failure,and even the failure to start,  the hero’s quest shows us that.

Promises, threats, and coalitions

As we add more players, all rationality may go out of the window as players decide whether to form coalitions or to perform strategic style voting. If we introduce the idea of the players communicating then we add the issues of trust in promises, or fear of threats and it all starts to sound rather Hunger Games.

On social media aggression and threats are common, because of prejudice, or group think, especially on Twitter where there is no moderation. And, online and off, we have all been promised things and relationships which have ultimately left us disappointed, and told us that we have been misinformed, like the fake news, we’ve been hearing about a lot lately.  Fake news is not new, in other contexts it is known as propaganda.  And,  if it is not completely fake, just exaggerated, well that’s new either, New Labour loved spin which led to a sexed up dossier, war and death.

Kermit’s next move

Philip D. Straffin says in his book Game theory and strategy, that game theory only works up to a point, after which a player must ask for some clarification about what is going on because mathematics applied to human behaviour will only explain so much.

And so we turn back to Kermit. What is he to do?  He has passive-aggressively asked for clarification and had a cup of tea. What’s his next move? Well, he could wait and see if he gets a reply (tit for tat). Who will crack first (chicken). But with the texts he has sent her, it is likely that her response is somewhat predetermined, or perhaps not, perhaps she will repond with Nash’s equilibria, or at the very least the Pareto principle of everyone not getting the worst outcome.

Alternatively, he could take a breath and remember that he is talking to someone he likes and with whom he wants to spend some time, someone human with the same vulnerabilities as him. He could adopt the volunteer’s dilemma approach and send her an honest text to explain that his feelings are hurt, he thought they had something special, and that she liked communicating with him as much as other people. By seeking clarification in this way, Kermit may just end up having a very nice evening after all –  or not. Whoever said: All’s fair in love and war, didn’t have instant access to social media and all the complications it can cause.

[Part 4]

Game theory in social media marketing (2): Customers and competitors

Source: rarewallpapers.com

[Part 2 of 4: Game theory & social media: Part 1, Part 3, Part 4]

In part 1, we saw how people love to play games. Game theory was first recognised in 1928, by John Von Neumann’s paper which was about two people playing a game together with only one winner (known as: two person game-zero sum).

If we apply game theory to social media marketing, we could say that the customer and the marketer are playing a two person game, zero sum – winner takes all. Before social media, this might have been the case, for customers believed that shops were acting in their own self interests and so they, the customer, did too. Everyone was out to get what they could. In reality though, the relationship is more of a win-win: Without the marketer, the customer might not learn about the product on offer and not buy or benefit from the product, and without the customer, the marketer doesn’t have a job at all.

Playing your customer

In his book, Social Media Marketing, Eric Anderson describes the marketer-customer as a two-way mutually dependent conflict and, points out that in the world of marketing everything is described combatively. There are marketing campaigns, killer apps and dead lists, which fit with game theory: Two parties with opposing and mutual interests both engaged in winning the outcome of combat.

For if the customer doesn’t engage and play the game then, they effectively kill the product, or even the market the product exists in. More worryingly for a marketer, if a customer engages and is an influencer, this customer with a few well placed tweets and reviews on a social computing site (their blog, Amazon, Goodreads) can begin a campaign which can sink a product. On his blog, Nathan Bransford describes how books have been effectively killed prior to publication due to bad reviews on Goodreads.

A nice equation given by Kyle Wong on Forbes describes what an influencer does as follows:

Influence = Audience Reach (# of followers) x Brand Affinity (expertise and credibility) x Strength of Relationship with Followers

Influencers have immense power to kill or create sales, which is a totally new thing in marketing. This is potentially such a powerful way to sell to millions across the globe, especially amongst certain demographics – mums, millennials –  that many companies view social media marketing as the only way to market nowadays. They know that they must, like influencers,  build relationships with their customers. One way to do this is by creating content.

Playing your competitor

In a great blog on coschedule.com, Julie Niedlinger, describes how game theory approaches to creating content can help marketers decide whether their strategy (another military word) is appropriate with the competitors and with their customers.

Niedlinger advises marketers to take a moment, before reacting to comments that potential customers will leave on blogs, in order to ask whether there is a game going on. If so which game? And most importantly, are the rules clear? Once they are then and only then should a marketer make a move.

Secondly, she looks at competitors producing a similar blog of content rich potentially market cornering information and asks what is the next move?  Do you steal their writers? Mimic them? Join forces? Or, follow trends in an effort to win their share of the market.

It is important to know your game, it’s rules, and the moves you should be taking.

In part 3, we will look at specific game theory theories and see what moves and games we could play.

[Part 3]

Game theory in social media (1): Fate and power

Early dice made from knucklebone in Ancient Greece © British Museum

[Part 1 of 4: Game theory & social media: Part 2, Part 3, Part 4]

Humans love games. We just love playing them.

The earliest proof we have, so far, dates back to 3600BCE: Six-faced dice with coloured pebbles made from heel bones of sheep and deer have been found on archaeological digs in Assyria, Sumeria, and Egypt.

Today, a walk around the British Museum, one of my favourite places, (which you can now do on Google), shows us that places where guards sat, probably for hours, like the entrance to the palace of King Ashurnasirpal II (883-859 BC) at Nimrud, has a board and a tally scratched into the side of one of the enormous winged human-headed lions.

By the time of the birth of Christ, many types of random number generators, including dice, were common, and were used for betting on or with board games. They were often spoken of as the workers of the blind goddess of fate, fortune, or destiny. And, it says in the Bible, that they cast lots to decide how to divide up Jesus’s possessions (Matthew 27:35). Even nowadays we talk about the roll of the dice when we talk chance and the things which happen to us.

By 10th century Europe, cards were the most popular thing with which to play games. There might be some skill, but really, a lot of it is up to chance, and don’t we all know that cliche about playing the hand you were dealt?

Highs and lows on the roll of a dice

The first formal attempt at analysing games, especially of chance, was written in 1520 (but published in 1663) by Gerolamo Cardano and has been recognised as the first step in probability theory. Cardano was a compulsive gambler, so would have felt the highs and lows of the roll of the dice more than most. He was foremost in the minds of Pascal and Fermat who published a book in 1654, continuing his work. And, it was Fermat’s last theorem which remained a phenomenon until it was solved in 1994. Imagine, it took three hundred and fifty years to solve a puzzle.

Later,  writer Fyodor Dostoyevsky described our love of excitement and chance when playing games and how our fortunes can flip in an instant. He wrote about it in letters to his sister and his short novel, The Gambler. He was convinced that you needed to detach and keep a clear head, but had difficulty doing either, for it is much easier said than done. Consequently, gambling and games are ubiquitous, from church bingo to nationwide lotteries. Life can really change with a roll of the dice – or so it seems.

Game theory not gamification

It was in 1928 that the first theory of game theory was first written about by (rock star) John von Neumann who amongst many things designed the first computer architecture in 1945.

But, it has to be said, game theory isn’t the same as gamification, at all. Please don’t mix them up. Gamification is about turning things into games such as business objectives and anything else we want to make more engaging and more fun. When gamification is well designed, it works really well. But game theory is much bigger, and much more than just games.

In 1944, von Neumann and Oscar Morgenstern translated and expanded von Neumann’s theories in order to produce: The theory of games and economic behaviour. For his 1928 paper was mainly about two people playing a game together with only one winner (known as: two person game-zero sum) but game theory is much bigger than this, and it is not just about games and game playing.

It might be based in mathematics, but game theory has people in it, of course, which is why it can be used to think about everything: economics, political science and psychology. And, it has the crazy assumption that people behave rationally, which if there is one thing I know about life, people never behave rationally, nor should you expect them to. The other thing is that, we can only partially model any prescription because the world is huge and constantly changing, and we can never model everything in a computer. It really doesn’t matter how clever computers get. We have a long way to go yet when modelling humans and behaviour, but game theory is a start.

That said, power is the name of the game: group voting, economic theory and how to influence people, especially in areas like interpersonal cooperation, competition,  conflict, labour negotiations, and economic duopolies, can all be understood in terms of game theory.

Game theory for explaining social media

Social media is the big new tool of the Internet, for business, politics, etc, and as of yet, no one knows how it works. So, this series is going to take a look at some of the big hitters of game theory: the prisoners dilemma, the Nash equilibrium, and so on, to see if these strategies can help us understand better how social media works. Are people cooperating or conflicting in ways these models describe on social media? If yes, can we understand and anticipate behaviour?  If not, what other theories could we come up with?

Let’s take a look.

[Part 2]

Designing design: Great expectations

project

[Part 8 of 12: 1) The science of the artificial 2) function, behaviour structure 3) form follows function, 4) no function in structure, 5) the medium is the message 6) types and schemas 7) aesthetics: attractive things work better 8) managing (great) expectations 9) colour 10) styles and standards 11) design solution spaces 12) conclusions]

Managing expectations is the key to success in most areas of life, not just design.

When a design artefact is judged to be useless, it is often because it does not behave in the way the user is expecting. This is because there is a gap between what the designer intended and what the user is expecting, which in user interface design, Donald Norman calls the gulf of execution. Straddling this gulf, is the way to manage user expectations.

Modelling

We all have models of how the world works (mental models), and of ourselves (self schemas) to explain and make sense of everything. In design, we have three models:

  • The designer’s model of how the artefact works.
  • The user’s model of how the artefact should work and how it actually works which changes with experience.
  • The artefact’s image of how it works  which is the way it looks which should be supported by the documentation or user manual.

These models should line up in order to match user expectations, but to do so, the designer has to provide the correct cues.

Cues

Affordance

First proposed by psychologist James Gibson, affordance describes how the physical properties of an artefact will influence its function. So, round wheels are much more suitable than square wheels on your bicycle. It is easy to see that the round wheels go round, square wheels might make you think the bicycle is an uncomfortable seat.

Consistency

When similar looking parts of an artefact or system work in a similar way then users can easily transfer what they have learnt from one part to another and have similar experiences. Consistency can be aesthetic, for example, the windows on a graphical user interface windows have the same layout, or the logo is the same on each restaurant chain outlet so the customer know what to expect. Consistency can also be functional such as how traffic lights work in a certain expected order: red, amber, flashing amber, green.

Constraints

Constraints are used to indicate what actions are possible. These can be physical such as when barriers are put up at sporting events to direct crowds and traffic. They may also be psychological in symbols such as a skull and cross bones for poison or danger, or conventional which we learn, such as we stop when the  traffic lights are red. Or, they can be cultural such as what people wear during mourning, in some countries it is black, in others it is white.

Feedback

Feedback is necessary to guide user behaviour. A dialog box can ask: Is this what you want to do? Less usefully it might say: No, you can’t do that.  The box really needs to add: but you can do this or this.

Mapping

We all need a link between what we do and what happens, so if you are driving your car and turn the steering wheel left you expect it to turn left.

Tradeoffs

In previous posts, we saw that the asethetically pleasing lemon squeezer known as Juicy Salif had sacrificed some of it functionality in order to look good, as did the Lockheed Lounger. In the same way, the more function you add to an artefact, particularly in the cues which are needed to guide and manage user expectations, the less usable it becomes. Complex gadgets may look cool but if they are not functional their value is more asthetic than usable and will only satisfy a tiny section of determined users.